OnBuy has deployed a new artificial intelligence solution, enabling its partners to unlock an estimated £4 million in gross merchandise value (GMV) in just six weeks.
The AI-powered auto-categorisation tool utilises a bespoke model paired with vector database technology to sort products listed on OnBuy's eCommerce marketplace across more than 6,600 available categories. According to the company, this deployment has more than doubled categorisation accuracy, with the latest figures showing an average accuracy rate exceeding 93%.
OnBuy stated that, to date, over 20 million products on the platform have been automatically categorised by the technology. Internal analysis estimates that performing this work manually would have required a team of five people more than two decades to complete, at a notional cost of over GBP £4 million.
The categorisation process is described as critical to the shopping experience, as it directly affects product discoverability and retailers' sales prospects. Senior staff noted that prior to the implementation, many products were incorrectly classified, making it more difficult for customers to find relevant items.
Following the technology rollout, OnBuy reported that 50,000 products originally listed prior to 2025 became newly discoverable through corrected categorisation, thereby contributing millions of pounds in previously untapped sales.
Chief Executive Officer and Founder, Cas Paton, said:
"Our latest AI-fuelled deployment highlights OnBuy's continued evolution as a platform. We're using this technology to transform our business, with only a fraction of the time and investment previously required. We've leveraged AI to increase categorisation accuracy, which means a better, more efficient shopping experience for customers and more sales for our retailers.
Paton also clarified the technical challenges his team overcame in developing this system. "There was no out of the box solution capable of handling our unique taxonomy and product data structure, but our team of leading tech experts were able to build an agent that could handle the task."
OnBuy reports that the technology is tailored to its marketplace's complex requirements, including compliance with regulatory rules such as age restrictions and value-added tax (VAT) obligations. The company affirmed that its solution is unique in facilitating large-scale auto categorisation while keeping regulatory complexities in mind.
Looking ahead, OnBuy plans to enhance the system to support product categorisation in non-English languages, intended to support future international expansion. The company also disclosed that further iterations of the software aim to simplify the onboarding process for new retailers joining the platform.
In the backdrop of this AI deployment, OnBuy has recently acquired new platforms to grow its market share. In May, the company announced an investment of GBP £10 million intended to revive the Comet brand as a new eCommerce destination within its expanding retail network.
OnBuy was founded in 2016 and operates as an online marketplace for verified retailers across a broad range of categories, such as electrical goods, toys, beauty products, and home and garden items. The platform's model is structured around supporting third-party retailers, without stocking its own inventory, and sharing cashback rewards with customers.
The company has received recognition in sector rankings, including listings among Europe's fastest-growing companies and the UK's fastest-growing private technology firms. Its operational aim is to increase sellers' exposure while providing consumers with greater choice and value.
OnBuy's leadership maintains that the integration of artificial intelligence will form a central pillar in the evolution of its marketplace platform, building further efficiencies and expanding possibilities for retail partners.
